inconsistent with the height of sixty or seventy miles, the least which has
hitherto been ascribed to it. This kind of Aurora is not brighter than the
milky way, and resembles sheet-lightning in its motions.
For the sake of perspicuity, I shall describe the several parts of the Aurora,
which I term beams, flashes, and arches. The beams are little conical pencils
of light, ranged in parallel lines, with their pointed extremities towards the
earth, generally in the direction of the dipping-needle. The flashes seem to
be scattered beams approaching nearer to the earth, because they are similarly
shaped and infinitely larger. I have called them flashes because their appearance
is sudden, and seldom continues long. When the Aurora first becomes
visible, it is formed like a rainbow, the light of which is faint, and the motion of
the beams undistinguishable. It is then in the horizon. As it approaches the
zenith, it resolves itself at intervals into beams, which, by a quick undulating
motion, project themselves into wreaths, afterwards fading away, and again
brightening, without any visible expansion or concentration of matter. Numerous
flashes attend in different parts of the sky. That this mass, from its short
distance above the earth, would appear like an arch to a person situated at the
horizon, may be demonstrated by the rules of perspective, supposing its parts
to be nearly equidistant from the earth. An undeniable proof of it, however,
is afforded by the observations of the 6th and 7th of April, when the Aurora
which filled the sky at Cumberland-House, from the northern horizon to the
zenith, with wreaths and flashes, assumed the shape of arches at some distance
to the southward.
But the Aurora does not always make its first appearance as an arch. It
sometimes rises from a confused mass of light in the east or west, and crosses
the sky towards the opposite point, exhibiting wreaths of beams, or coronas
boreales, in its way. An arch, also, which is pale and uniform at the horizon,
passes the zenith without displaying any irregularity or additional brilliancy ;
and we have seen three arches together, very near the -northern horizon, one of
which exhibited beams and even colours, but the other two were faint and
uniform.
On the 7th -of April, an arch was visible to the southward, exactly similar to
that in the north, and it disappeared in fifteen minutes. It had probably passed
the zenith before sun-set. The motion of the whole -body of Aurora is from the
northward to the southward, -at angles not more than 20° from the magnetic
meridian. The centres of the arches were-as often in the magnetic as in the
true meridian.
The colours do not seem to depend on the presence of any luminary, but to
be generated by the motion of the beams, and then only when that motion is
rapid, and the light brilliant. The lower extremities quiver with a fiery red
colour, and the upper with orange. We once saw violet in the former. The
number of Aurorae visible in September was two; in October three ; in November
three; in December five; in January five; in February seven; in
March sixteen; in April fifteen; and in May eleven. Calm and clear weather
was the most favourable for observation; but it is discernible in cloudy weather,
and through mists. We could not perceive that it affected the weather.
The magnetic needle, in the open air, was disturbed by the Aurora, whenever
it approached the zenith. Its motion was not vibratory, as. observed by Mr.
Dalton; and this was, perhaps, owing to the weight of the card attached to
it. - It moved slowly to the E. or W. of the magnetic meridian, and seldom
recovered its original direction in less than eight or nine hours. The greatest
extent of its aberration was 45'.
A delicate electrometer, suspended at the height of fifty feet from the ground,
was never perceptibly affected by the Aurora, nor could we distinguish its
rustling noise, of which, however, such strong testimony has been given to us,
that no doubt can remain of the fact. The conclusions to be drawn from the
above will be found in the observations for the winter of 1820.
An Account of the Aurora Borealis, seen at Cumberland-House, between the 23d
of October, 1819, and the 13tli o f June, 1820; Lat. 53° 56' 40" N. Long,
102° 16' 41" W. Variation 17° 17' 31" E.
From the 23d of October to the 25th of November, the Aurora was not
visible, or it did not appear before one a. m.
November 26th, at one a. m., an Aurora, arched like a rainbow, about 20°
high; centre bearing north ; colour pale yellow, faint. At eight p.m., a very
faint arch, centre north.
December 6th, at ten p.m., a faint-arched Aurora, centre N. b. E.
8th. A similar Aurora, centre north, at ten p.m.